Sensemaking With/About AI: Unpacking Design Professionals' Data Sensemaking Styles in a High-Stakes Industrial Context

Designers face opportunities and challenges in leveraging AI to make sense of increasingly available data. There is limited research on how designers engage with AI in data sensemaking to inform design decisions, especially in safety-critical contexts. To address this knowledge gap, we co-designed and evaluated an AI-supported prototype that supports data exploration using truck interaction design as a case study. Our analysis highlighted a bidirectional relationship between AI and data, where designers used AI to gain an overview of user scenarios, and used data to contextualize and validate AI outputs. We identified three sensemaking styles, illustrating how designers flexibly orchestrated expertise, data, and interaction to interpret AI outputs. Our findings contribute to an in-depth characterization of AI-supported data sensemaking and offer implications for designing tools that empower diverse, situated, and critical engagement with data through AI in high-stakes settings.